SEAT Ibiza - rwd turbo
Discussion
This week, more Coding on the cars powertrain electronics:
Turns out there was a bit of a f**kup in some of the code running in the throttle control ecu, which was occasionally blocking transmission requests, oops! As the code is written to try to ensure no "unintended acceleration" the ratification and arbitration of external throttle requests is complex, especially when those requests exceed the current driver demand. Anyway, problem fixed (after spending about a day going through code i wrote 5 years ago and have almost completely forgotten since......)
Then, yet more coding, starting to get my embedded ABS controller up and running on a 32b platform:
It's a tough life at the front eh?? ;-)
Turns out there was a bit of a f**kup in some of the code running in the throttle control ecu, which was occasionally blocking transmission requests, oops! As the code is written to try to ensure no "unintended acceleration" the ratification and arbitration of external throttle requests is complex, especially when those requests exceed the current driver demand. Anyway, problem fixed (after spending about a day going through code i wrote 5 years ago and have almost completely forgotten since......)
Then, yet more coding, starting to get my embedded ABS controller up and running on a 32b platform:
It's a tough life at the front eh?? ;-)
Edited by anonymous-user on Thursday 29th May 16:25
Fastdruid said:
I'd never have guessed that board was custom.
Nice work.
LOL, i think that was the last "through hole" device i designed, in about 2006 iirc (although it went through a few revisions, rather than the re-design it should have got, over the next few years. Since then, i'm a SMC man, and enjoy the massive space saving benefits, if not the eye strainingly small component packages ;-)Nice work.
Like this:
pretty much twice the capability of that old ecu, in around 4x less space.........
Even nicer! 
I'm designing some boards for my car and while initially planning on through hole have decided to 'bite the bullet' and go SMT. My first experience of soldering SMD by hand was in making the Megasquirt and wasn't _too_ bad although I'm not sure about the IC's!
Who do you get to make your boards? I was thinking of going to Seeed Studios ( http://www.seeedstudio.com/service/index.php?r=sit... )
I'm designing some boards for my car and while initially planning on through hole have decided to 'bite the bullet' and go SMT. My first experience of soldering SMD by hand was in making the Megasquirt and wasn't _too_ bad although I'm not sure about the IC's!
Who do you get to make your boards? I was thinking of going to Seeed Studios ( http://www.seeedstudio.com/service/index.php?r=sit... )
Fastdruid said:
Even nicer!
I'm designing some boards for my car and while initially planning on through hole have decided to 'bite the bullet' and go SMT. My first experience of soldering SMD by hand was in making the Megasquirt and wasn't _too_ bad although I'm not sure about the IC's!
Who do you get to make your boards? I was thinking of going to Seeed Studios ( http://www.seeedstudio.com/service/index.php?r=sit... )
Get yourself a cheapo hot air gun (the ebay 858D from Atten amongst others are cheap, but usable), and if you want to do big stuff (ie big power silicon mounted on massive ground planes) also get either a hot plate or use a DIY hot air gun (carefully) to heat the entire pcb up a bit first! With decent "fine" solder paste, some good tweasers and ideally some form of magnifier, SMC really isn't too hard, and because you don't have to snip /bend leads etc it's faster two. The downsides are there is significantly less "bodge room" on the pcb, so your design needs to be spot on, and your component footprints are critical, unlike for TH stuff where you have some nice long lead wires to er, adjust, if your spacing is a bit wonky etc!I'm designing some boards for my car and while initially planning on through hole have decided to 'bite the bullet' and go SMT. My first experience of soldering SMD by hand was in making the Megasquirt and wasn't _too_ bad although I'm not sure about the IC's!
Who do you get to make your boards? I was thinking of going to Seeed Studios ( http://www.seeedstudio.com/service/index.php?r=sit... )
I get most of my PCBs from Techbridge Circuits, who are effectively a UK agent for various chinese board houses. They are expensive (relatively) but provide a super efficient easy to use service. They also check your designed before they send them, which helps avoid stupid mistakes / over sights etc. For a beginner, Henry who runs it is super helpful on the phone, and that really helps you get your manufacturing data well sorted, which the first time round can be confusing. For high volume production on a proven board, i'd go straight to the "online" chinese board houses, but for prototype stuff where you just need it to be right, Techbridge is my choice ;-)
Here's a Techbridge pcb, with the gold finish instead of the normal (cheaper!) HotAirSolderLevelled (HASL) finish. The quality of the pcb, and the screen print/solder masking is excellent, and in probably over 1000 pcbs of various designs i've never had a board with a manufacturing defect (they 100% electrical test iirc)
;-)
;-)
!amusingduck said:
Max_Torque said:
This week, more Coding on the cars powertrain electronics:
What language is it coded in?
I've been playing around with the powersteering system for a while now, trying to get a reasonably nice, linear handwheel weighting across the widest range of handwheel velocities, and trying to avoid having hydraulic lag or "catch up" weight up the wheel just when you need to stick in some rapid inputs during a bit of a "moment" etc!
Due to space constraints with the tight engine packaging, i use a TRW Electrohydraulic PAS pump, "borrowed" from a Ford Focus ST. This runs into my GrpA Cossie steering rack (only 1.2 turns Lock to Lock).
Due to the fast over all steering ratio, the rack requires quite a high fluid flow rate, and i was getting significant "Lag" and a corresponding un-acceptable increase in handwheel loading.
The TRW pump takes vehicle and engine speed and handwheel rate over CAN, and changes it's pump speed to suit, and although i have had this working, i've never been happy with the way it responds in my car, with a steering system quite different from the original ST.
Wondering if i am using the pump to it's full potential, i disassembled my spare unit, and "sucked out" the calibration data from the EEPROM on the control PCB:
Plotted graphically this is what you get for the 512bytes of data:
It's pretty obvious that there are multiple "pump speed" maps in there! (in fact, 6 of them)
Using some educated guessing, it's possible to work out what the X and Y axis are, and build the maps up, with each one looking something like this:
Unfortunately, at this time, i had no idea how to get the unit to switch between running those maps. ie, what the CAN command was to change it's operating mode.
Step forwards generous PH member Podie who bravely offered up his rather orange ST as a guinea pig! Logging the CAN data on his car, whilst we switched between drive modes (comfort/normal/sport) and wiggled the handwheel / drive around the block, quickly furnished me with about 30mb of coms bus data to wade through! And success, i now have a pump that can be selected between it's internal assistance maps!
I use my "Chassis dynamics module" to send data to the pump, meaning i can filter or modify the true (measured) vehicle speed and handwheel rate, allowing me to remap the assistance characteristic to suit my chassis.
As the data is all shoved over CAN it's easy to log and see whats going on:
This has massively improved the steering capability, although my next trick going to be a slight increase in the output volume of the pump by modifying the gearset and housing, with a reduction in the peak pressure relief valve setting. In conjunction with reducing the pump motor current shunt resistance by around 20% i think i can get a significantly better pump match to my steering rack and system ;-)
Due to space constraints with the tight engine packaging, i use a TRW Electrohydraulic PAS pump, "borrowed" from a Ford Focus ST. This runs into my GrpA Cossie steering rack (only 1.2 turns Lock to Lock).
Due to the fast over all steering ratio, the rack requires quite a high fluid flow rate, and i was getting significant "Lag" and a corresponding un-acceptable increase in handwheel loading.
The TRW pump takes vehicle and engine speed and handwheel rate over CAN, and changes it's pump speed to suit, and although i have had this working, i've never been happy with the way it responds in my car, with a steering system quite different from the original ST.
Wondering if i am using the pump to it's full potential, i disassembled my spare unit, and "sucked out" the calibration data from the EEPROM on the control PCB:
Plotted graphically this is what you get for the 512bytes of data:
It's pretty obvious that there are multiple "pump speed" maps in there! (in fact, 6 of them)
Using some educated guessing, it's possible to work out what the X and Y axis are, and build the maps up, with each one looking something like this:
Unfortunately, at this time, i had no idea how to get the unit to switch between running those maps. ie, what the CAN command was to change it's operating mode.
Step forwards generous PH member Podie who bravely offered up his rather orange ST as a guinea pig! Logging the CAN data on his car, whilst we switched between drive modes (comfort/normal/sport) and wiggled the handwheel / drive around the block, quickly furnished me with about 30mb of coms bus data to wade through! And success, i now have a pump that can be selected between it's internal assistance maps!
I use my "Chassis dynamics module" to send data to the pump, meaning i can filter or modify the true (measured) vehicle speed and handwheel rate, allowing me to remap the assistance characteristic to suit my chassis.
As the data is all shoved over CAN it's easy to log and see whats going on:
This has massively improved the steering capability, although my next trick going to be a slight increase in the output volume of the pump by modifying the gearset and housing, with a reduction in the peak pressure relief valve setting. In conjunction with reducing the pump motor current shunt resistance by around 20% i think i can get a significantly better pump match to my steering rack and system ;-)
Firstly, really nice to meet you yesterday.
Incredible amount of data goes into these things - quite staggering to realise just how much work goes into a power steering pump!
Glad you've got the data you need, but should you need more, you know where to find me and the the citrus Focus
Incredible amount of data goes into these things - quite staggering to realise just how much work goes into a power steering pump!
Glad you've got the data you need, but should you need more, you know where to find me and the the citrus Focus
Firstly, really nice to meet you yesterday.
Incredible amount of data goes into these things - quite staggering to realise just how much work goes into a power steering pump!
Glad you've got the data you need, but should you need more, you know where to find me and the the citrus Focus
Incredible amount of data goes into these things - quite staggering to realise just how much work goes into a power steering pump!
Glad you've got the data you need, but should you need more, you know where to find me and the the citrus Focus
Gassing Station | Readers' Cars | Top of Page | What's New | My Stuff